CSE 250 B Assignment 1 Report

نویسندگان

  • Yuncong Chen
  • Pengfei Chen
  • Yang Liu
چکیده

In this report we analyzes two types of logistic regression models and carries out experiments using the UCI Adult dataset. We compare several factors that related to the prediction accuracy and discuss reasons of those correlations. Finally we propose a reasonable configuration of parameters and schemes for training models. Experiments show that for both models we achieved around 84% accuracy.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

CSE 250 B Assignment 4 Report

In this project, we implemented the recursive autoencoder (RAE) as described in Socher’s paper to discover the sentiment of sentences. We train and test our RAEs with a dataset of over 10000 sentences from movie reviews, and achieve 75.4% accuracy.

متن کامل

CSE 250 B Assignment 3 Report

Latent Dirichlet Allocation (LDA) is a probabilistic, generative model designed to discover latent topics in text corpora, and it can be learned by collapsed Gibbs sampling. In this report, we evaluate the effectiveness of LDA by experiments on two dataset, Classic400 and BBC. We discuss related issues in Gibbs sampling, including goodness-of-fit criteria, parameter tuning, convergence, etc., a...

متن کامل

Anodal Transcranial Pulsed Current Stimulation: The Effects of Pulse Duration on Corticospinal Excitability

The aim is to investigate the effects of pulse duration (PD) on the modulatory effects of transcranial pulsed current (tPCS) on corticospinal excitability (CSE). CSE of the dominant primary motor cortex (M1) of right first dorsal interosseous muscle was assessed by motor evoked potentials, before, immediately, 10, 20 and 30 minutes after application of five experimental conditions: 1) anodal tr...

متن کامل

CSE 255: Assignment 1 - Exploring Musical Tagging

We explore two predictive tasks: (i) a measure of tag probability, and (ii) identifying a minimum tag set for more meaningful music classification on a 100,000 song dataset joined across complementary databases from the 1 Million Song Dataset (“MSD”). We conclude that a tag set size of around 50 tags is most meaningful and report many of our findings/analysis based on the top 50 tags. Using lin...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2012